NAVER Cloud GPU Instance Features
NAVER Cloud Platform is a major cloud service provider in South Korea, widely utilized by enterprise, financial, and public sector organizations. Based on local data center infrastructure, it offers low network latency, KRW billing, and standard Korean tax invoice compliance.
Its core GPU lineup includes enterprise accelerators like the NVIDIA A100 and V100, optimized for localized workloads and AI training pipelines.
On-Demand Pricing and Billing Structure
NAVER Cloud publishes instance rates on an hourly and monthly basis. Keep the following factors in mind:
- VAT Exclusion: Published prices exclude 10% VAT, meaning final billing statements reflect a 1.1x multiplier.
- Per GPU-Hour Calculation: Multi-GPU instances (e.g., 8-GPU configurations) are billed as complete server units. Calculating the exact per GPU-hour rate is essential when comparing with global GPU cloud providers.
- Egress and Storage Fees: Outbound data transfer and high-performance block storage carry separate charges, which can add up for inference serving.
Commitment Discounts and Savings
For sustained production serving or long-term AI research, 1-year to 3-year commitment plans provide discounts over on-demand rates. Korean startups can also apply domestic AI voucher credits to offset infrastructure expenses.
How to Check on Perhour
Visit the A100 price page to compare NAVER Cloud's per GPU-hour rate directly against other domestic and international cloud providers with tax-excluded normalization.